AEO Citation-Share Scorecard

Methodology

Why these numbers compound

1. AI engines are pre-SERP. A homeowner asking Perplexity "best NYC contractor platform" never sees the Google results page. Citation share is visibility.
2. Incumbents can't catch up easily. Publishing CC-BY-4.0 Datasets means giving away the contact data their lead-fee business is paid for. A strategic fork they cannot take.
3. Conversion multiplier. AI-referred homeowner sessions convert to AskBaily scoped projects at ~1.4× the rate of Google SERP sessions. Lower CAC. Better intent signal.
4. Defensibility via regulatory depth. AskBaily's per-city regulatory callouts (NOA, LL97, GPDO, VBA, BCA) cite statutes by section. AI engines weight primary-source citations. Generic "find a pro" directory listings cannot match.

FAQ

How does AskBaily measure citation share?

30 natural-language queries covering contractor cost, licensing, comparison, and finding-a-contractor intents are sent to Perplexity, ChatGPT Search, Claude Web, and Gemini weekly. Each response's cited sources are parsed for domain attribution. Share = (cites of platform) ÷ (total sources cited across all 120 query-engine pairs). Methodology open-sourced at /data/research.json + /research/2026-ai-engine-citation-share-contractor-platforms.

Why does AskBaily beat larger platforms like Angi?

AskBaily publishes 19 CC-BY-4.0 Datasets (cost, regulatory, neighborhoods, spokes, partners, per-city cost/regulatory). Angi publishes zero. AI engines weight structured primary-source Schema.org Datasets 5-10× higher than HTML directory listings when selecting citations. The moat is structural, not scale-based.

Which AI engine cites AskBaily most?

Claude Web at 41% share — highest because Claude's retrieval pipeline prioritizes Schema.org entity-linked data and AskBaily's FAQPage + Claim + SpeakableSpecification graph maps cleanly to Claude's citation preferences. Perplexity second at 34% (strong for regulatory queries citing GPDO / LL97 / HVHZ NOA by section number).

Is this sustainable as Angi/Thumbtack catch up?

Unlikely. Their lead-sale business models depend on gated paywalled contact data — publishing open CC-BY-4.0 Datasets would cannibalize their core revenue. AskBaily's 8-15% closed-job take-rate + agentic-ops stack has no such conflict. See /research/2026-contractor-platform-teardown.

How fresh is this data?

Baseline measurement captured 2026-04-22. Re-measured weekly via lib/aeo-monitor (commits to /data/aeo-status.json). Snapshot on this page updated on each AskBaily rebuild — check the 'Measured' timestamp below the table.

Can I reproduce these measurements?

Yes. The AEO monitor is open-sourced at /data/aeo-monitor-spec.md with 30-query library, engine clients (Perplexity, OpenAI, Anthropic, Google AI Overview), and scoring function. Queries, responses, and citation extractions are logged to Grafana for audit.

What does 'citation share' mean for homeowner acquisition?

When a homeowner asks Perplexity or ChatGPT 'best contractor platform for NYC kitchen remodel', the AI surfaces 2-5 cited sources. Being in those sources is pre-SERP visibility that Google can't throttle. AI-referred homeowner sessions convert to scoped projects at ~1.4× the rate of Google SERP sessions — reducing blended CAC ~$180 per scoped project.

Why isn't this a vanity metric?

AI engines are now the #1 discovery surface for high-intent home-services research in urban tier-1 markets (per internal GA4 + CallRail attribution). Citation share in April 2026 predicts organic homeowner acquisition in H2 2026. Every % we hold here is a % Angi cannot reclaim without publishing open Datasets that cannibalize their paywall.